The present invention relates to antibodies and antibody fragments which bind to specific conformational epitopes of P-selectin, and to methods of their use and identification.
In normal hemostasis and immune surveillance, leukocytes circulate freely in the blood and respond to injury and infection in a sequential process of adhesion signaling mediated by cell adhesion molecules (1-3). In inflammatory and thrombotic disease, this process is dysregulated and can sustain pathology wherein leukocytes attack the body's own tissue and can cause serious and sometimes deadly complications. It is well known that leukocyte adhesion plays a major role in the pathology of many inflammatory and thrombotic disorders such as vasoocclusion in sickle cell disease, reperfusion injury, thrombosis, atherosclerosis, asthma, rheumatoid arthritis, psoriasis and tumor metastasis (4-15) deep venous thrombosis (DVT). P-selectin is also involved in other disease processes, such as tissue and organ damage associated with inflammation, e.g., ischemia-reperfusion injury. P-selectin is thus a target for intervention in human inflammatory and thrombotic diseases.
Selectins are a family of adhesion proteins which are known to play key roles in the recruitment of leukocytes to activated endothelium and platelets. P-selectin is a member of the selectin family of adhesion glycoproteins which also includes L- and E-selectins (16). The selectins mediate the recruitment, initial tethering and rolling, and adherence of leukocytes to sites of inflammation (1). P-selectin is stored in Weibel-Palade bodies of endothelial cells and alpha-granules of platelets and is rapidly mobilized to the plasma membrane upon stimulation by vasoactive substances such as histamine and thrombin (17).
Sickle Cell Disease
Sickle cell disease (SCD) is a rare inherited blood disorder that causes chronic anemia and vasoocclusion, affecting primarily people of African-American heritage in the United States. Sickle cell disease is the most common single gene disorder in African Americans, affecting approximately 1 in 375-600 persons of African ancestry (18, 19). Sickle cell conditions are also common among people of Mediterranean countries, Africa the Caribbean and parts of South and Central America (18, 19).
SCD is an autosomal recessive disease caused by a single missense mutation (Val6Ala) in the β-globin gene that renders the mutant hemoglobin less soluble and prone to polymerization upon deoxygenation. The polymerization of hemoglobin causes deformation of the erythrocyte to give the cell a sickled shape (20).
SCD has three common variants: homozygous sickle cell disease (hemoglobin SS disease), doubly heterozygous sickle hemoglobin C disease (hemoglobin SC disease) and the sickle β-thalassemias. The most common and severe form of the disease occurs in individuals who inherit two copies of the HbS variant (HbSS) and the primary hemoglobin in their red blood cells is sickle hemoglobin. Other individuals can be affected as compound heterozygotes with varying severity of the disease. They have one copy of the HbS variant paired with a copy of another β-globin gene variant. HBSC results in a mild form of the disease. Hb β-thalassemia variants (resulting in the inability to produce the normal βA globin chain (β°) or a reduction in its production (β+) result in a range of clinical severities. HbS β° is a severe form, whereas HbS β+ can be moderate or mild based on the contribution of each variant to the total hemoglobin of the patient. Other more rare variants can result if in conjunction with the S gene, another abnormal hemoglobin is inherited from the other parent, such as D, G or O. The predominant form of sickle cell is present in individuals with one copy of HbS and one copy of the normal β-globin gene (HbA). These individuals carry the sickle cell trait (18).
SCD affects an estimated 50-100,000 people in the US (21-24). All individuals that are homozygous or compound heterozygous for HbS show some clinical manifestations of SCD. Symptoms usually appear within the first 6 months of life but there is considerable variability in SCD severity (25). Individuals with HbSS are most severely affected, followed by individuals with HBbS β°-thalassemia (22, 26). In addition to genotype, additional factors affect disease severity such as the levels of fetal hemoglobin and the haplotye of the β-globin cluster, a region that contains 5 genes that code for the β portion of hemoglobin. Despite the capacity to determine genetic risk factors, the ability to predict disease course from birth is limited (27).
In the USA, sickle cell screening at birth is mandated in all 50 states and the District of Columbia (28) and offers an opportunity for early intervention. Diagnostic testing methodology usually comprises a complete blood count in conjunction with one or more of hemoglobin electrophoresis, isoelectric focusing, high-performance liquid chromatography and DNA testing (22).
Chronic Anemia and Hemolysis
The sickled erythrocyte has a shorter half-life than the normal erythrocyte and results from the instability of HbS and the effects of repeated episodes of hemoglobin polymerization/depolymerization in the circulation. This affects membrane ionic permeability, cellular viscosity and deformability (20) and promotes oxidative membrane damage (29). Sickle cell disease patients are anemic by 2 to 3 months of age and develop symptoms and complications associated with chronic anemia and hemolysis (22, 30) such as renal disease, ophthalmic disorders, leg ulcers, priapism and pulmonary hypertension (26, 31-37). Hemoglobin values for SCD patients range from 6 to 10 g/dL and the hemoglobin S molecule has a poor affinity for oxygen. Triggers for transfusion in patients are a hemoglobin value of 5 or less or a precipitous drop in hemoglobin of 2 g/dL or more. Transfusions are typically given to restore hemoglobin values to baseline levels established for each patient as excessive hemotocrit can precipitate sickling (38). SCD patients are more susceptible to parvovirus B19 infection which can arrest erythropoiesis and lead to aplastic anemia crisis (39).
Vasoocclusive Pain Crisis
Vascular occlusion is central to the clinical course of SCD and likely involves both the micro and macro circulation. Occlusion occurring in the microvasculature can culminate in acute painful episodes or vasoocclusive pain crises. Vasoocclusive pain crisis is the clinical hallmark of microvascular occlusions and accounts for over 90% of hospital admissions of adults SCD patients. It is well known that polymerization of hemoglobin S during deoxygenation and cell sickling leads to blockage of the microvasculature (40). However, it has recently become clear that hemoglobin S polymerization is not solely responsible for vasoocclusion. It has now been demonstrated that such events as sickled red cell lysis, cell membrane damage and oxidative stress, repeated ischemic damage, and microvasculature injury due to the adhesive interactions between sickle red cells and the endothelium that culminate in a proinflammatory environment (41-43). In this environment of chronic vascular inflammation, the adherence of leukocytes, platelets and sickled red cells to activated blood vessel endothelium and to each other is believed to be a primary cause of microvasculature blockage and vasoocclusive pain crisis (43-47). Additional factors such as the rigidity of sickled cells, increased blood viscosity, and local vasoconstriction have also been identified as potentially contributing to the vasoocclusion process.
Long-term repeated vasoocclusive events and occlusions occurring in the macrovasculature can cause life-threatening complications leading to organ damage and failure, stroke and death (40). There is an approximately 20 to 30 year reduction in life expectancy in sickle cell disease patients (48). Chronic pain in SCD is not just a continuation of the pain of vasoocclusion: it is usually secondary to avascular necrosis of bone at various joints (49). Sickled red cells can become trapped in the spleen causing it to become enlarged and precipitating splenic sequestration crisis causing sudden and severe anemia. Functional asplenia leaves patients susceptible to infection (18). Bone growth retardation, renal (32), ophthalmic (33) and cerebrovascular complications (ranging from clinically evident acute stroke to transient silent ischemic infarct) (50) are seen as major clinical consequences of SCD and vasoocculsive injury (22). Acute chest syndrome is another major complication (51), and is a significant cause of morbidity and mortality (52).
Pain episodes appear to be triggered by a number of factors including cold, stress and physical exertion (38, 53). The frequency, severity, location and duration of pain crises can vary considerably, even within a specific disease subtype. Patients with homozygous sickle cell and sickle cell β°-thalassemia have a higher frequency of vasoocclusive pain crises than patients with hemoglobin SC and sickle cell-β°-thalassemia genotype (54). Disease severity is thought to depend on a complex interaction of genetic, rheologic and hematologic factors, as well as microvascular and endothelial factors. Crises commonly involve pain in the back, legs, knees, arms, chest and abdomen (53). The frequency of crisis and pain severity varies considerably among patients and in the same patient over time. One study evaluating pain rates in patients ranging from newborns to age 50 years indicated that 5.2 percent of patients with sickle cell disease have three to 10 episodes of severe pain every year (54). In an independent study, over 30% of sickle cell patients in the US (approximately 27,000 patients) have three or more pain crises per year (55). Moreover, a recent study (PISCES) evaluating health related quality of life issues in SCD patients indicated that pain crisis might be significantly underreported among SCD patients (56).
Current Therapies for Vascular Occlusion
Vascular occlusion in SCD patients can manifest in multiple ways including vasoocclusive pain crisis, acute chest syndrome, cerebrovascular events and multiple types of organ failure. Therefore, treatment modalities for vascular occlusion depend on the clinical course and severity of the disease and are generally symptomatic or palliative in nature. Patient education in the avoidance of initiating factors that precipitate vasoocclusive pain crisis has shown some prophylactic benefit. The two most common symptomatic treatments are blood transfusions and analgesics. Most SCD patients commonly have hemoglobin values between 6 and 10 g/dL and hemoglobin values typically drop at least 1 g per dL during a vasoocclusive pain crisis. Severe pain resulting from vasoocclusive crisis can be treated with narcotics but their use is controversial due to concerns of narcotic addiction and tolerance. Other complications with narcotic use are drug-seeking behavior, sedation and respiratory depression. Oxygen management has been utilized to treat vasoocclusive pain crisis despite the lack of strong evidence supporting its effectiveness. Rehydration is also sometime used during vasoocclusive pain crises with some benefit (22, 38).
Bone marrow transplantation may be considered and can be curative, but its use is restricted to a limited number of patients, and carries a high risk of morbidity and mortality (22).
Hydroxyurea (Droxia) is the only FDA approved drug for treatment of SCD pain crises. The mechanisms by which it produces its beneficial effects are uncertain but may involve increasing hemoglobin F levels in RBCs thereby decreasing the level of hemoglobin S polymerization. Hydroxyurea is cytotoxic, myelosuppressive and teratogenic (57, 58) which implies a carcinogenic risk to SCD patients. The long-term effects however, on hematologic toxicities, organ damage and carcinogenicity are currently unknown (59, 60).
In summary, most therapies for vasoocclusive pain crisis in SCD patients provide symptomatic relief and do not address the underlying cause of this debilitating condition. To date only one therapy has been approved by the FDA for the treatment of pain crisis, thus, patients with SCD represent a major unmet medical need in a life-threatening disease with severe morbidities.
P-Selectin as a Therapeutic Target for SCD
In SCD, as noted above, interactions between sickled red cells, platelets, leukocytes and the microvasculature are P-selectin-dependent processes and result in vasoocclusion and painful crisis. Studies in transgenic mice engineered to express human β hemoglobin S (βS) have shown that antibody-mediated inhibition of P-selectin function can prevent and/or reduce vasoocclusion, indicating a therapeutic potential for this target. In addition mice expressing the βS hemoglobin that lack P-selectin (due to gene deletion) do not suffer vasoocclusion, further supporting a key role for this molecule in this morbidity.
The hyper-inflammatory state in SCD patients is characterized by activated monocytes and vascular endothelium (61-63). A similar proinflammatory phenotype was demonstrated in resting state βS mice which exhibit elevated levels of peripheral leukocytes and neutrophils, an increased number of rolling and adherent leukocytes, and reduced blood flow volume and red blood cell velocities (64). The βS mice were hypersensitive to hypoxia/reoxygenation resulting in an inflammatory response represented by a significant increase in the number of adherent and emigrated leukocytes. This inflammatory response was completely blocked by a functionally blocking anti-mouse P-selectin antibody, but not by a functionally blocking anti-mouse E-selectin antibody, demonstrating a critical role for P-selectin in this process.
Inflammatory Bowel Disease
Inflammatory Bowel Disease (“IBD”) is the collective term used to describe two chronic, idiopathic inflammatory diseases of the gastrointestinal tract: ulcerative colitis (“UC”) and Crohn's Disease (“CD”). UC and CD are considered together because of their overlapping clinical, etiologic, and pathogenetic features. From a therapeutic and prognostic standpoint, however, it is useful to distinguish them.
IBD occurs world-wide and is reported to afflict as many as two million people. Onset has been documented at all ages; however, IBD predominately begins in young adulthood. The three most common presenting symptoms of IBD are diarrhea, abdominal pain, and fever. The diarrhea may range from mild to severe and is often accompanied by urgency and frequency. In UC, the diarrhea is usually bloody and may contain mucus and purulent matter as well. Anemia and weight loss are additional common signs of IBD. Reports of an increasing occurrence of psychological problems, including anxiety and depression, are perhaps not surprising secondary effects of what is often a debilitating disease that occurs in people in the prime of life.
A battery of laboratory, radiological, and endoscopic evaluations are combined to derive a diagnosis of IBD and to assess the extent and severity of the disease. Nevertheless, differentiating UC from CD, as well as other types of inflammatory conditions of the intestines, such as irritable bowel syndrome, infectious diarrhea, rectal bleeding, radiation colitis, and the like, is difficult, because the mucosa of the small and large intestines reacts in a similar way to a large number of different insults. Once other types of bowel disorders have been ruled out, the final diagnosis is often made on the basis of the progression of the disease. In many patients, though, the colitis must still be regarded as indeterminate because of the overlapping features of UC and CD, particularly with CD of the colon.
The leading early symptoms of UC and CD are chronic recurrent diarrhea, bloody diarrhea, recurrent abdominal pain, nausea, weight loss general evidence of inflammation without any obvious explanation (fever, raised ESR, leucocytosis, thrombocytosis and dysproteinenemia or anemia). Among these symptoms, diarrhea and anemia are more characteristic of UC while pain and weight loss and marked evidence of inflammation are more common in CD. While the history and physical examination of a patient can help, the final confirmation of the diagnosis has traditionally been made endoscopically, histologically and, in relation to the small intestine, radiologically as well.
The SAMP-1/Yit mouse model of spontaneous iletis closely resembles human Crohn's Disease (65, 66). Therapeutic inhibition of PSGL-1 binding uniquely ameliorates ileitis in this model whereas blockade of individual selectins does not (67). Inhibition of TNF in this model does reduce the severity of ileitis in a manner similar to anti-PSGL-1 binding although the therapeutic effect does not appear to be as potent as anti-PSGL-1. Thus, the SAMP-1 model appears to closely mirror human Crohn's Disease not only in its pathophysiology but also in its response to therapeutic intervention. This evidence points to the conclusion that therapeutic substances which inhibit P-selectin-PSGL-1 binding activity in humans (and other primates) would also be effective as treatments of Crohn's Disease.
P-selectin plays its central role in the recruitment of leukocytes to inflammatory and thrombotic sites by binding to its counter-receptor, P-selectin glycoprotein ligand-1 (PSGL-1) (or a PSGL-1-like receptor on sickled red blood cells), which is a mucin-like glycoprotein constitutively expressed on leukocytes including neutrophils, monocytes, platelets, and on some endothelial cells (68). The ultimate physiologic function of the selectins is to promote extravasation of leukocytes into inflamed or damaged tissues. The initial binding of P-selectin on the endothelium to PSGL-1 on the leukocytes is essential and central to this process. The predominant mechanism for rolling and tethering of leukocytes to activated endothelium and platelets is the binding of leukocyte PSGL-1 to the P-selectin on these cells (68, 69). PSGL-1 binds to P-, L- and E-selectin (70). P-selectin and SGP-3, a glycosulfopeptide modeled from the N-terminus of PSGL-1, have been co-crystallized and the contact residues for lectin-ligand binding have been identified (71).
The selectins share common structural motifs including a lectin domain (or carbohydrate recognition domain), an epidermal growth factor-like domain (EGF), a varying series of consensus repeats, a transmembrane domain and a cytoplasmic tail (70). As noted, the initial tethering and rolling of leukocytes is mediated by the interaction of P-selectin and PSGL-1. Thus the blocking of P-selectin function by using (1) antibodies to P-selectin, (2) antibodies to PSGL-1, (3) fragments of PSGL-1 or recombinant forms of PSGL-1, (4) small molecules that mimic the binding domain of PSGL-1, and (5) other molecules that disrupt the binding of P-selectin to PSGL-1, can block leukocyte rolling and tethering and thus prevent firm adhesion to endothelial cells or platelets. Mice deficient in P-selectin or PSGL-1 also fail to support leukocyte tethering and rolling on activated endothelial cells (72, 74). L-selectin plays a dual role in that it is constitutively expressed on circulating leukocytes and can initiate “secondary binding” by interaction with PSGL-1 on other leukocytes (75). This process leads to further recruitment of new leukocytes to the inflamed area. L-selectin binding to PSGL-1 also plays a role in homing of lymphocytes to the high endothelial vasculature (HEV) venules in the secondary lymphatic system (76). E-selectin is transcriptionally regulated and is expressed on activated endothelial cells hours after P-selectin mediated events. E-selectin can bind PSGL-1 with low affinity but can also bind other ligands. Single transgenic knockout mice for each selectin have shown that these molecules possess compensatory selectin mechanisms for leukocyte homing and rolling (77).
In view of the above, there is a well-established need for new treatments, such as antibodies, that target P-selectin as a means of treating inflammatory and thrombotic diseases by disrupting the binding of P-selectin and PSGL-1. It is therefore a preferred goal of the present invention to block P-selectin binding to PSGL-1 to block the adherence of blood cells that contribute to vasoocclusion in SCD and other thrombotic disorders.